Search

Amyloid β oligomers (Aβo) are the main toxic species in Alzheimer's disease, which have been targeted for single drug treatment with very little success. In this work we report a new approach for identifying functional Aβo binding compounds. A tailored library of 971 fluorine containing compounds was selected by a computational method, developed to generate molecular diversity. These compounds were screened for Aβo binding by a combined 19F and STD NMR technique. Six hits were evaluated in three parallel biochemical and functional assays. Two compounds disrupted Aβo binding to its receptor PrPC in HEK293 cells. They reduced the pFyn levels triggered by Aβo treatment in neuroprogenitor cells derived from human induced pluripotent stem cells (hiPSC). Inhibitory effects on pTau production in cortical neurons derived from hiPSC were also observed. These drug-like compounds connect three of the pillars in Alzheimer's disease pathology, i.e. prion, Aβ and Tau, affecting three different pathways through specific binding to Aβo and are, indeed, promising candidates for further development.

To identify small molecules that shield mammalian sensory hair cells from the ototoxic side effects of aminoglycoside antibiotics we screened 10,240 compounds, selecting those that protected against neomycin and gentamicin in zebrafish lateral-line hair cells and, when retested in mouse cochlear cultures, prevented gentamicin-induced death of outer hair cells (OHCs). Of 64 compounds that protected zebrafish hair cells, 8 protect OHCs from gentamicin in vitro. These hits share structural features and all block, to varying degrees, the OHC’s mechano-electrical transducer (MET) channel, a known route of aminoglycoside entry into hair cells. Further characterisation of one of the strongest MET-channel blockers, UoS-7692, revealed it additionally protects against kanamycin and tobramycin, and does not abrogate the bactericidal activity of gentamicin. UoS-7692 behaves, like the aminoglycosides, as a permeant blocker of the MET channel, significantly reduces gentamicin-Texas Red loading into OHCs, and preserves lateral-line function in neomycin-treated zebrafish. Trans-tympanic injection of UoS-7692 protects mouse OHCs from furosemide-kanamycin exposure in vivo and partially preserves hearing. The results confirm the hair-cell MET channel as a viable target for the identification of compounds that protect the cochlea from aminoglycosides, and provide a series of hit compounds that will inform the design of future otoprotectants.

Lead optimization (LO) in medicinal chemistry is largely driven by hypotheses and depends on the ingenuity, experience, and intuition of medicinal chemists, focusing on the key question of which compound should be made next. It is essentially impossible to predict whether an LO project might ultimately be successful, and it is also very difficult to estimate when a sufficient number of compounds has been evaluated to judge the odds of a project. Given the subjective nature of LO decisions and the inherent optimism of project teams, very few attempts have been made to systematically evaluate project progression. Herein, we introduce a computational framework to follow the evolution of structure-activity relationship (SAR) information over a time course. The approach is based on the use of SAR matrix data structures as a diagnostic tool and enables graphical analysis of SAR redundancy and project progression. This framework should help the process of making decisions in close-in analogue work.

A systematic analysis of ionization states of current bioactive compounds is presented. Ionization states were related to biological activities on the basis of high-confidence activity data. The majority of bioactive compounds were found to be basic or neutral under physiological conditions. In addition, chemical neighborhoods of active compounds frequently contained analogs with different ionization states that were activity-conservative. However, a variety of targets were identified that displayed clear preferences for specific ionization states in compounds active against them. In this context, notable differences in the distribution of ionization states were detected for compounds active against different target superfamilies. Furthermore, under physiological pH, differences in ionization states of active compounds were tolerated by many targets. However, in a number of instances, ionization states of highly and weakly potent compounds active against the same target were found to be distinct, providing guidelines for compound design and optimization.

Activity cliffs (ACs) are defined as pairs of structurally similar compounds sharing the same biological activity but having a large difference in potency. Therefore, ACs are often studied to rationalize structure-activity relationships (SARs) and aid in lead optimization. Hence, the AC concept plays an important role in compound development. For compound optimization, ligand efficiency (LE) represents another key concept. LE accounts for the relation between compound potency and mass. A major goal of lead optimization is to increase potency and also LE. Despite their high relevance for drug development, the AC and LE concepts have thus far not been considered in combination. It is currently unknown how compounds forming ACs might be related in terms of LE. To explore this question, ACs were systematically identified on the basis of high-confidence activity data and LE values for cliff partners were determined. Surprisingly, a significant increase in LE was generally detected for highly potent cliff partners compared to their lowly potent counterparts, regardless of the compound classes and their targets. Hence, ACs reveal chemical modifications that determine SARs and improve LE. These findings further increase the attractiveness of AC information for compound optimization and development.

Matched molecular pairs (MMPs) are defined as pairs of compounds that only differ by a chemical change at a single site. MMPs have become popular in medicinal chemistry to support lead optimization, absorption, distribution, metabolism, excretion, and toxicity (ADMET) analysis, and other applications. Thus far, MMPs have been algorithmically defined and not on the basis of reaction information. This often limits the chemical interpretability and practical utility of MMPs. Therefore, we introduce synthetically accessible MMPs that are automatically generated by applying reaction rules following the retrosynthetic combinatorial analysis procedure (RECAP). A library of more than 92 000 RECAP-MMPs was generated from public domain compounds active against 435 different targets exclusively utilizing high-confidence activity data. This library is made freely available for use in medicinal chemistry.

We have searched for chemical transformations that improve drug development-relevant properties within a given class of active compounds, regardless of the compounds they are applied to. For different compound data sets, varying numbers of frequently occurring data set-dependent transformations were identified that consistently induced favorable changes of selected molecular properties. Sequences of compound pairs representing such transformations were determined that formed pathways leading from unfavorable to favorable regions of property space. Data set-dependent transformations were then applied to predict a series of compounds with increasingly favorable property values. By database searching the desired biological activity was detected for several designed molecules or compounds that were very similar to these molecules. Taken together our findings indicate that data set-dependent transformations can be applied to predict compounds that map to favorable regions of molecular property space and retain their biological activity.

Activity cliffs (ACs) are formed by structurally similar compounds with large differences in activity. Accordingly, ACs are of high interest for the exploration of structure-activity relationships (SARs). ACs reveal small chemical modifications that result in profound biological effects. The ability to foresee such small chemical changes with significant biological consequences would represent a major advance for drug design. Nevertheless, only few attempts have been made so far to predict whether a pair of analogues is likely to represent an AC-and even fewer went further to quantitatively predict how "deep" a cliff might be. This might be due to the fact that such predictions must focus on compound pairs. Matched molecular pairs (MMPs), defined as pairs of structural analogs that are only distinguished by a chemical modification at a single site, are a preferred representation of ACs. Herein, we report new strategies for AC prediction that are based upon two different approaches: (i) condensed graphs of reactions, which were originally introduced for modeling of chemical reactions and were here adapted to encode MMPs, and, (ii) plain descriptor recombination-a strategy used for quantitative structure-property relationship (QSPR) modeling of nonadditive mixtures (MQSPR). By applying these concepts, ACs were encoded as single descriptor vectors used as input for support vector machine (SVM) classification and support vector regression (SVR), yielding accurate predictions of AC status (i.e., cliff vs noncliff) and potency differences, respectively. The latter were predicted in a compound order-sensitive manner returning the signed value of expected potency differences between AC compounds.

Macrocyclic compounds experience increasing interest in drug discovery. It is often thought that these large and chemically complex molecules provide promising candidates to address difficult targets and interfere with protein–protein interactions. From a computational viewpoint, these molecules are difficult to treat. For example, flexible docking of macrocyclic compounds is hindered by the limited ability of current docking approaches to optimize conformations of extended ring systems for pose prediction. Herein, we report predictions of bioactive conformations of macrocycles using conformational search and binding modes using docking. Conformational ensembles generated using specialized search technique of about 70 % of the tested macrocycles contained accurate bioactive conformations. However, these conformations were difficult to identify on the basis of conformational energies. Moreover, docking calculations with limited ligand flexibility starting from individual low energy conformations rarely yielded highly accurate binding modes. In about 40 % of the test cases, binding modes were approximated with reasonable accuracy. However, when conformational ensembles were subjected to rigid body docking, an increase in meaningful binding mode predictions to more than 50 % of the test cases was observed. Electrostatic effects did not contribute to these predictions in a positive or negative manner. Rather, achieving shape complementarity at macrocycle-target interfaces was a decisive factor. In summary, a combined computational protocol using pre-computed conformational ensembles of macrocycles as a starting point for docking shows promise in modeling binding modes of macrocyclic compounds.

Matched molecular pairs (MMPs) consist of pairs of compounds that are transformed into each other by a substructure exchange. If MMPs are formed by compounds sharing the same biological activity, they encode a potency change. If the potency difference between MMP compounds is very small, the substructure exchange (chemical transformation) encodes a bioisosteric replacement; if the difference is very large, the transformation encodes an activity cliff. For a given compound activity class, MMPs comprehensively capture existing structural relationships and represent a spectrum of potency changes for structurally analogous compounds. We have aimed to predict potency changes encoded by MMPs. This prediction task principally differs from conventional quantitative structure-activity relationship (QSAR) analysis. For the prediction of MMP-associated potency changes, we introduce direction-dependent MMPs and combine MMP analysis with support vector regression (SVR) modeling. Combinations of newly designed kernel functions and fingerprint descriptors are explored. The resulting SVR models yield accurate predictions of MMP-encoded potency changes for many different data sets. Shared key structure context is found to contribute critically to prediction accuracy. SVR models reach higher performance than random forest (RF) and MMP-based averaging calculations carried out as controls. A comparison of SVR with kernel ridge regression indicates that prediction accuracy has largely been a consequence of kernel characteristics rather than SVR optimization details.